No announcement yet.

reliable HS version with mcsSprinklers

  • Filter
  • Time
  • Show
Clear All
new posts

  • reliable HS version with mcsSprinklers

    Which version of HS (2.0.x or 2.1.x) currently gives the most reliable operation with mcsSprinklers? I plan to switch to mcsSprinklers as I can't get the Rain8Net plug-in to work reliably under 2.0.x or 2.1.x. The Rain8Net plug-in (or maybe it's HS or a MS problem?) suddenly stops being able to turn on valves anywhere from 2-14 days after it was initially started.

    I intend to only use the Rain8Net hardware for valve control (no weather station, sensors, etc.)

    I'm looking for a specific verison number, such as 2.0.2061.


  • #2
    I have never tried it under HS2.1, but I suspect others have and clearly HS2.0 is near its end of life.

    There are two other options with Rain8Net that are available and reduce dependence on HS operations, but still allow HS to be involved. One is the standalone version of mcsSprinklers and the other is xapmcsRain8Net. In both of these cases the mcsXap plugin to Homeseer can be used tol form a LAN interface to allow HS device status and control. It is possible to run either of these independent of HS, but when HS is up and running it can synchronize to provide a central status location for all your HS activites.

    With xapmcsRain8Net there is a scripting interface so you could develop your own control logic or your could use Homeseer events and scripting to drive it. The interface logic for this software was taken from mcsSprinklers so the core elements are mature, but it is a relatively new posting as an independent package.

    With mcsSprinklers the control logic is built-in via a browser interface. The core logic for mcsSprinklers is the same between the standalone and HS plugin.

    xapmcsRain8Net is at
    standalone mcsSprinklers is at


    • #3
      I just installed HS 2.0.2041 and mcsSprinklers. I am able to control all of my Rain8Net valves.

      When I set a certain area to be "interval" for 2 days, the zone status shows that the next watering will occur 4 days from now (Thursday) instead of Tuesday. When I set the interval for 1 day, it correctly shows the watering will occur tomorrow (Monday). I did reset the last watering time to now, too. Is this a bug or do I have something else set incorrectly?


      • #4
        Do you have any restrictions setup that prohit it from running on Tuesday? If not then I need to look at what may be causing the incorrect prediction. The prediction is informative so if there is no other reason for it not running on Tuesday then it will run even if the prediction is wrong.


        • #5
          I haven't entered any restrictions in the restrictions setup. I just did a "save changes" on the restrictions setup hoping it would initialize something that was omitted, but that didn't make a difference.

          If I set it to certain days of the week, then it does show the next day/time correctly.


          • #6
            Something is definitely wrong with my mcsSprinklers configuration. Today an area was watered that should not have been watered.

            The area (#2) that was watered was programmed for every 2 days (interval). The zone status screen said this should happen at 9:00am. Instead, it was watered immediately following the watering of another area (#1) that I initiated manually around 6:30am. This other area (#1) has each zone configured as "disable automatic" control, so that's why I manually started it.

            I have all automatic irrigation disabled until I determine what is wrong.

            Michael, how can I troubleshoot this?


            • #7
              When I try to force various intervals I do not have any problems with the actual scheduling. I did find the problem with the predicted scheduling that moved the date out an extra two days. To debug the sceduling problem the best route is to enable the General Debug on the Other page. It will produce a file in the \Data\mcsSprinklers folder.

              I recently went to IE for download of internet data and I'm now finding that it becomes unstable when using IE and am now working to remove it and replace it with the Inet control. IE also has a big footprint that tend to slow down the startup initialization.

              While I wanted to do more testing before posting, I did post V2.6.29 that does contain what I believe to fix the interval prediction and it also removes IE.
              Last edited by Michael McSharry; May 3rd, 2006, 01:07 AM.


              • #8

                I just installed version 2.6.29. In case you are interested, I did not have the msinet.ocx file in my \WINDOWS\System32 directory (XP Pro SP2), so I moved your file there.

                The predicted times are correct now. I will be watching the log very frequently to see if it behaves as it should.

                Since the IE method of retrieving wx info was unreliable, does this mean I should check the "Get Using API" radio button instead of the "Get Using IE" button? Wouldn't it make sense to remove the "Get Using IE" option from the setup entirely?

                How do I disable all internet weather information retrieval? I'd like to do that while this portion is still being proven by others.


                • #9
                  No provisions now exist to disconnect, but it is a good idea since there will be some who want the timed scheduling and don't care about internet data or have a dial up and don't want to use the bandwith for this application. Right now if you do not select any AWS sites then the internet access will be only one access per every 30 to 60 minutes.

                  I'm going to leave both methods available to get data since different PC configurations behave differently and I want to provide some flexibilty to accomodate a wider range of setups. I am actually implementing a second way to deal with the API interface since many users will not have Windows Scripting 5.6 installed and I want to have a mechanism to monitor and kill spawned internet access processes that do not complete for whatever reason.

                  I should have this out soon. There are a few other enhancements now in work that need to complete testing first.